The Study On Key Technologies And Micro-vision System For Flexible Substrate Detection

Flexible substrate is a FPC-based package board. The future trend of chip package is to use flexible substrate in high-density chip package. Currently, the flexible substrate has been developed to the micron level. However, the traditional manual inspection and automatic detection system cannot automatically detect the main parameters of this kind of flexible substrate. To solve this problem, a micro-vision system is built for flexible substrate detection based on microscopic detection method, while some key technologies are studied as well. The main aspects of the research are as follows:(1) According to the characteristics of high-density ultra-thin flexible substrate, a micro-vision system is set up for flexible substrate detection. Both hardware selection and software module design are completed.(2) To meet the real-time requirements of the system, ‘Z’ shaped image acquisition method is used. In addition, multi-task parallel processing method is designed to improve system efficiency based on the analysis of system tasks relationship.(3) Image distortion, which is caused by the microscope objective lens and camera lens, is analyzed under nonlinear camera model. On this basis, image distortion is corrected based on variable ratio method and Tsai method.(4) In order to obtain overall information of the flexible substrate, characteristics of detecting system and the flexible substrate microscopic image are analyzed. To stitch images, three methods are compared, including phase correlation method, block matching method based on motion estimation and mixing ratio method based on correlation factor. Based on the comparisons, a logarithmic difference method and a fast multi-resolution image stitching method are proposed to stitch flexible substrate microscopic images according to the characteristics of these substrates. Furthermore, weighted average method is applied to fuse microscopic images. Experimental results show that the usage of the fast multi-resolution image stitching method and the weighted average method results in high speed and high accuracy in image mosaic and fusion.(5) To provide prior information for the microscopic image stitching, coarse location method of microscopic images is designed according to the movement regularity of the system. Finally, based on the conversion model between the overall flexible substrate image and the standard image, the microscopic images are then located on the standard image according to the results of microscopic image stitching.Throughout this thesis, designs, methods and key technologies of detection system for high-density ultra-thin flexible substrate manufacturing process are provided.